Institutionen för fysik, kemi och biologi

Examensarbete

Microcontact printing of antibodies in complex with

conjugated polyelectrolytes

Fredrik von Post

Examensarbetet utfört vid forskningsgruppen för biomolekulär och

organisk elektronik (Biorgel)

2007-10-22

LITH-IFM-EX--07/1872—SE

Linköpings universitet Institutionen för fysik, kemi och biologi

581 83 Linköping

Institutionen för fysik, kemi och biologi

Microcontact printing of antibodies in complex with

conjugated polyelectrolytes

Fredrik von Post

Examensarbetet utfört vid forskningsgruppen för biomolekulär och

organisk elektronik (Biorgel)

2007-10-22

Handledare

Anna Herland

Jens Wigenius

Examinator

Olle Inganäs

Table of contents

1 Abbreviations.............................................................................................................. 1 2 Introduction................................................................................................................. 3

2.1 Aim ..................................................................................................................... 4 3 Experimental techniques............................................................................................. 5

3.1 Microcontact printing.......................................................................................... 5 3.1.1 PDMS.......................................................................................................... 5 3.1.2 Replica molding and microcontact printing................................................ 5

3.2 Printing preparations........................................................................................... 6 3.2.1 Conjugated polyelectrolytes........................................................................ 7 3.2.2 Proteins and antibodies ............................................................................... 7

3.3 Luminescence and fluorescence ......................................................................... 7 3.4 Conjugated polyelectrolytes................................................................................ 9 3.5 Fluorescence spectroscopy................................................................................ 11 3.6 Fluorescence microscopy.................................................................................. 12 3.7 Imaging ellipsometry ........................................................................................ 13 3.8 AFM.................................................................................................................. 17

4 Results and discussions............................................................................................. 19 4.1 Fluorescence spectroscopy................................................................................ 19

4.1.1 Comparison of protein interaction between four different conjugated polyelectrolytes ......................................................................................................... 19 4.1.2 Interaction of POWT and aHSA............................................................... 22

4.2 Fluorescence spectroscopy – discussion........................................................... 23 4.3 Microcontact printing........................................................................................ 24

4.3.1 Microcontact printing of antibody ............................................................ 24 4.3.2 Microcontact printed antibody in complex with CPE............................... 29 4.3.3 Microcontact printed antibody interacting with antigen........................... 29 4.3.4 Microcontact printed antigen interacting with antibody........................... 30 4.3.5 Printed antibody interacting with antibodies in solution .......................... 30

4.4 Microcontact printing - discussion.................................................................... 35 5 Conclusions............................................................................................................... 37 6 Recommendations..................................................................................................... 39 7 Acknowledgements................................................................................................... 41 8 References................................................................................................................. 43

1

1 Abbreviations

AFM atomic force microscopy Alexa Alexa 488 AMCA amino methylcoumarin CPE conjugated polyelectrolyte EtOH ethanol FITC fluorescein isothiocyanate GABA γ-amino butyric acid aGABA γ-amino butyric acid antbody HSA human serum albumin aHSA human serum albumin antibody IgG immunoglobulin gamma µcp microcontact printing PBS phosphate buffered saline PDMS poly(dimetylslioxane) POMT poly (3-[(s)-5-amino-5-methoxylcarboxyl-3-oxapentyl]-2,5-

thiophenylenehydrochloride) POWT poly (3- [(s)-5-amino-5-carboxyl-3-oxapentyl]-2,5-

thiophenylenehydrochloride) PTAA polytiophene acetic acid PTT poly (3-[2,5,8-trioxanonyl] thiophene) TxR Texas red

2

3

2 Introduction

Patterning is of great importance in many areas of modern science and technology, the biggest application of patterning is probably production of integrated circuits, information storage devices and display units. The underlying notion of patterning in technological applications is to manufacture devices with high information density. The smaller the features of the pattern, the more information can be held within a given area. The process of patterning is commonly referred to as lithography, a flow of information which typically begins with a pattern design in the form of a dataset and ends at a patterned surface of a substrate.

Whitesides et.al presented an approach to patterning suitable for biological applications called soft lithography[1], including a number of patterning methods which all shared the common feature of using a elastomeric stamp for generating patterns and structures. One of these techniques is called microcontact printing (µcp), it is used in this work and the main steps involved are outlined below.

In a typical procedure (see figure 1) the steps are as follows a) fabricating of a topographically patterned master template, b) molding this template with a soft polymer to generate a patterned stamp c) the patterned face of the rubber stamp is covered with a molecular solution, dried and d) brought into conformal contact with the surface of the substrate. As in the case with a common office stamp, where the ink is transferred to the paper, the molecular species are transferred from the stamp to the substrate thereby reproducing the pattern. Depending on the intended application the ink used varies; examples are proteins including antibodies [2], and alkanethiols on gold [3]

Master fabrication is usually done using conventional lithographic tools, with the costs associated to those techniques. However, numerous stamps can be made from one master template and the same stamp can be used to replicate a pattern many times thus making microfabrication on a small scale possible with small means.

A biological assay typically comprises a series of operations to check for the presence or activity of a biologically relevant analyte in a liquid sample. There are many advantages linked to having the assay take place at a solid surface. Separation of bound and free species is easily done by washing the surface and thereby concentrating the available sensor signal to a two-dimensional layer minimizing the use of reagents. The ability to use many different recognition molecules on well-defined locations on a surface enables effective screening of complex solutions of biological analyte. This is exemplified by the high resolution DNA array and the launch of the industry putting this technique to work [4].

Even though microcontact printing of antibodies and subsequent detection of antigen has been showed[5, 6], much work in the area of microcontact printing of antibodies has used secondary antibodies to evaluate the functionality and the likeness to template pattern of the printed antibodies [7]. A perhaps even more straightforward approach is to use the corresponding antigen as an indicator of the printed antibody pattern. The advantage of using the antigen as a discriminator is obvious; not only is the geometric features of the pattern evaluated but more importantly is it possible to evaluate the specific binding ability of the physiosorbed antibodies. This is the approach most used in this work.

4

One notion regarding this approach is that it is a very straightforward and perhaps not the most likely to succeed. It is in many ways an extreme and punishing environment for the delicate structures of the molecules involved. This might be an explanation to the few reports of successful applications of the technique.

2.1 Aim

The purpose of this diploma work is to evaluate the method of microcontact printing of functional antibodies for specific detection of complementary antigens by using conjugated polyelectrolytes designed and synthesized by Linköping University as reporter molecules.

A possible application of the detection technique described in this work was proposed by Ola Hermansson, Department of Neuroscience at Karolinska Institutet. The idea is to detect neurotransmitter substances exerted by neural stemcells growing on hard substrates. By using some modified version of the protein chips described, it could be possible to combine a growth substrate with the ability to detect neural cell specific substances. Such substances might be γ-amino butyric acid (GABA) or dopamine.

5

3 Experimental techniques

3.1 Microcontact printing

3.1.1 PDMS

Poly(dimethylsiloxan) (PDMS) is a material increasingly used for pattern replication and microfluidic systems. PDMS gives the possibility of fast manufacturing of microstructures from a silicon template without the need of a cleanroom environment. The PDMS used in this project was of the brand Sylgard 184 from Dow Corning Corporation. Sylgard 184 is distributed as a two component system, consisting of base and curing agent. These are to be mixed as one part curing agent together with ten parts base. PDMS will cure in room temperature although the curing is hastened in elevated temperatures.

The reasons behind the increasing usage of PDMS and soft lithography in favour of more traditional lithography are several, the most important being the easy handling of the material and the possibility of rapid manufacturing of microstructures[8]. But there are also several intrinsic properties which renders it suitable.

The transparency of PDMS is a great benefit to the evaluation of experiments. All kinds of optic microscopy, including fluorescence microscopy, can be used without any additions to the methods. PDMS is also gas permeable. This allows gaseous reaction products from fluidic channels to dissipate through the material. It also enables filling of channels with only one opening.

Another important property is the elasticity of the material. This is a good feature as well as probably the biggest drawback for creating submicrometer structures. It is useful as a cured PDMS cast can cling to curved or non-smooth surfaces. The elastic PDMS will conform somewhat to the material and adhere in spite of unevenness of the surface. This property also enables easy release of casts from moulds.

Work has shown that structures in Sylgard 184 smaller than 1 µm tend to merge or collapse during inking and printing. Better accuracy to sub-micrometer layouts can be achieved with a stamp polymer having a higher elastic modulus and greater surface hardness. In order to maintain the elasticity of the material, however, the modulus and surface hardness can not be too high. This leads to brittle materials that can no longer adapt and conform to a substrate [9].

3.1.2 Replica molding and microcontact printing

Casting of PDMS on a surface followed by curing in an oven for a sufficiently period of time gives a submicron resolution of the surface structures. This is an easy method for rapid replication of the wafer surface without any obvious deformation. The casting procedure can be repeated without damaging the template.

A master template is first created using conventional photolithographic methods. The templates used in this work were created by other workers at IFM. The creation of the master template is outlined below.

6

1. Computer assisted design of pattern.

2. Creation of shadow mask i.e. the design is transferred to a chromium pattern on a UV-susceptible quartz plate using a dedicated mask printer apparatus.

3. A silicon wafer is coated with a UV-sensitive substance (photoresist).

4. As UV-light is shined through the mask onto the photoresist layer certain areas will be shadowed by the mask and other areas will be exposed, the resist will react differently to exposure and non-exposure and a pattern is created.

5. The template wafer is treated with a silane solution to facilitate replica moulding of PDMS.

Replica molding with PDMS is a simple procedure. The first step is mixing of PDMS by taking one part curing agent and ten parts base. This is mixed for a few minutes and then put in a degassing chamber for removing trapped air. When the material is free from bubbles it is simply poured over the silicon wafer. The wafer is placed in an oven for curing. After curing the cast is peeled of from the wafer. This is usually not a critical step, and PDMS easily releases from the silanised silicon mould. When dealing with very thick PDMS slabs there is a somewhat increased risk for damaging of the structure at the release.

Figure 1, Outline of microcontact printing.

3.2 Printing preparations

The rubber cast from the master template was cut into smaller pieces, (50-100 mm2), using a scalpel. The pieces were kept with the patterned face up in polystyrene Petri dishes. Stamps were immersed in 50% EtOH and ultrasonically washed for 5 minutes prior to applying the inking solution.

7

The substrates used were standard object glass slides used for fluorescence microscopy, silicon wafers (Esbo, Finland) used for AFM- and ellipsometric imaging and silicon wafers with a thicker ~1000 Å silicon dioxide layer (Okmetic, LMTD Esbo, Finland) used for fluorescence microscopy and AFM imaging.

Substrates were treated to the TL1-washing procedure (washed with a 5:1:1 mixture double distilled water 18 mΩ (Milli-Q water), 30% hydrogen-peroxide (H2O2) and 25% ammonia (NH3) for 10 minutes in 80ºC) or in some cases cleaned in 65% NH3-acid for 24h and then rinsed in flowing deionized water.

In some cases the rubber stamps were treated with RF-air plasma in a plasma chamber (Pico-RF from Diener electronic, Germany) to produce a hydrophilic surface before inking. The stamp surfaces were exposed to the air plasma (150 W) for 10 seconds at 10 mBar.

3.2.1 Conjugated polyelectrolytes

POWT, POMT, PTT and PTAA have been synthesised at the Chemical Department, IFM, Linköping University. A polyelectrolyte chain has approximately a molecular weight of 3125 Da, giving a concentration of 320-360 µM for 1 mg/ml (the uncertainty due to the variation of the length of the polymeric chains). A stock solution of the polyelectrolytes was prepared, where the polyelectrolytes were diluted in Milli-Q water. POWT was diluted to the concentration of 0.5 mg/ml. PTAA, PTT and POMT had the concentration of 1 mg/ml. The stock solutions were stored at 6ºC and shielded from light. The polyelectrolytes were ultrasonicated for 5 minutes and diluted from these stock solutions to the final concentration in the corresponding buffer solutions prior to the experiments.

3.2.2 Proteins and antibodies

Human serum albumin (HSA) and GABA were purchased from Sigma (St. Louis, USA). Anti-HSA (150 kDa, 10 g/l), anti-GABA (150 kDa, 0.6 g/l) and anti-IgG (150 kDa, 7.9 g/l) were all purchased from DakoCytomation (Glostrup, Denmark). Anti-streptavidin labelled with Texas-red (aStreptavidin-TR) was prepared in a stock solution at the concentration of 2.55 mg/ml in Milli-Q water and streptavidin labeled with aminomethylcoumarin (Streptavidin-AMCA) and fluoresceinisothiocyanate (Streptavidin-FITC) was dissolved to the concentration of 0.5 mg/ml in 1:1 part of Milli-Q water and glycerol, all purchased from Rockland Inc. (Gilbertsville, USA).

3.3 Luminescence and fluorescence

Fluorescence is a molecular phenomenon in which a substance absorbs light (excitation) and subsequently radiates light of longer wavelength (emission). The event is very fast, on the order of nanoseconds. A special case of is phosphorescence, where emission is much slower and the substance is seen to emit light after the excitation light is extinguished. [10]

8

The mechanism of fluorescence is described by electronic excitation and relaxation. The electrons in a molecule are, according to quantum theory, only allowed certain fixed energy levels or states. Transfer of an electron to a higher energy state requires external energy and energy is also transferred from the molecule when electrons “fall” from higher states. The energy needed to excite an electron from a lower state to a higher is provided by photon quanta. Incoming radiation is absorbed by the molecule if the photon energy is matched by the difference of two permitted energy levels.

Figure 2, Morse plot of excitation and relaxation

These energy levels are called singlet states. A number of vibrational energy levels are associated to every singlet state. Electronic transition between vibrational energy levels takes place without emission or absorption of photons, the energy is equilibrated by transferring into vibration, rotation or translation of surrounding molecules [10] Transitions between singlet states on the other hand can be accompanied by the emission of a photon, giving rise to fluorescence. As some of the incoming photon energy is lost as heat the emitted photons are always of lower energy than the absorbed photons. This can be seen when comparing absorption spectrum for a molecule system with the corresponding emission spectrum. The energy shift between the two spectra is called the Stoke’s shift. The shape of absorption and emission spectra is related to the probability corresponding to specific electronic transitions.

9

The quantum mechanical version of the Franck-Condon principle states that the molecule undergoes a transition to the upper vibrational state that most closely resembles the vibrational wavefunction of the vibrational ground state of the lower electronic state. According to the Franck-Condon principle, the nuclear framework of an excited molecule remains constant during electronic transition. The most intense (most probable) vibronic transition is from the ground state to the vibrational state lying vertically above it in a Morse potential diagram i.e. same internuclear separation, as shown in Figure 2 [10]

This being said it is obvious that spectroscopy can give us detailed information about the energetic state of a molecule. Conformational changes, with the twisting and turning of nucleus and electronic systems will certainly affect the overlap integral determining transition probability. Changes in spectroscopic characteristics provides a tool for measuring conformational changes in a conjugated polyelectrolyte[11]

3.4 Conjugated polyelectrolytes

Conjugated polymers have a framework of alternating single and double carbon–carbon bonds. Single bonds are referred to as σ-bonds, and double bonds contain a σ-bond and a π-bond. All conjugated polymers have a σ-bond backbone of planar, overlapping sp2 hybrid orbitals. The remaining out-of-plane pz orbitals on the carbon atoms overlap with neighboring pz orbitals to give π-bonds[12]

Although the chemical structures of these materials are represented by alternating single and double bonds, in reality, the electrons that constitute the π-bonds are delocalized over the entire molecule (See Figure 3). Delocalized electrons are a prerequisite for conductivity of polymers. The behavior of conjugated polymers is dramatically altered with chemical doping and like inorganic semiconductors this increase their conductivity extremely [13]

Figure 3, polythiophene, conjugated polymer with alternating single and double bonds

Conjugated polymers with an electron removed from the valence band (p-doped) polymers have wide application—for example, electrochromic devices, rechargeable batteries, capacitors, membranes, charge dissipation, and electromagnetic shielding. Doping of the polymers also give rise to new luminescence properties. Conjugated polymers are of increasing importance in research of more effective LED devices and solar cells.[12]

Let’s consider the polymer molecules used as reporter molecules in some instances of this work.

10

Figure 4, poly(3-[(S)-5-amino-5-carboxyl-3-oxapentyl]-2, 5-thiophenylene hydrochloride), POWT

POWT is a conjugated polymer made water soluble by adding an ionic substituent (the amino acid serine) to the polymeric backbone. These polymers are called conjugated polyelectrolytes, CPEs. The charge of the zwitterionic amino acid changes with buffer system and forces the polymer backbone to adapt different conformations and thus changing the spectroscopic properties of the POWT molecule. This change which geometrically is described as going from a planar to more non-planar conformation is called an intra-chain process. It has been shown that the proximity of polymeric backbones from different POWT molecules is another major contributor in determining the fluorescent properties. [14]. Polymer proximity is generally prohibited by the presence of large biomolecules e.g. immunoglobulins, i.e. if the polymer chains are associated to a larger molecule the distance between them will increase and aggregation is less common. This is followed by an increase in fluorescence activity compared to aggregated solutions as it reduces the diffusion of the excited state to places where radiationless transitions can occur[15]. This is called an intra-chain process.

Non-covalent electrostatic interactions between the amino acid sidechain and biomolecules in solution give rise to a peak shift in a spectroscopic emission curve through the intra- and inter-chain processes described above. It has been shown that the conformational changes of biomolecules can be governed using non-covalent association of luminescent polyelectrolytes as conformational changes of biomolecules can lead to different conformations of the polymer backbone leading to an alteration of the absorption and emission properties from the polymer[11, 16, 17]. These conformational changes in the biomolecules are often induced by the presence of a second biomolecules, e.g. complementary DNA strands or enzyme cofactors, recognizing the first biomolecules. This makes the polyelectrolyte-biomolecule complex in effect functioning as a reporter molecule for the second biomolecule. This technique shows good potential for being used as a new sensor technology.

11

One of the most diverse biological recognition processes is the interaction between an antibody molecule and its specific antigen. The vertebraic humoral immune system is capable of producing specific antibodies to virtually any kind of molecule. In binding to the antigen the chemical micro environment of the antibody is dramatically altered. It is believed that one or many polyelectrolyte chains associated to the antibody molecule will experience these changes and respond by adapting the polymeric backbone to a more energetically favored state thus changing the effective conjugation length. Another effect is that the distance between different polymeric chains increases. This change in the electronic structure is then monitored by spectroscopy.

3.5 Fluorescence spectroscopy

Spectroscopy is the science of analyzing absorbance and emission of a substance over a spectrum of wavelengths. This is done by keeping the exciting wavelength fixed and recording the absorbtion/emission spectrum or recording the absorbtion/emission at a fixed wavelength for a spectrum of exciting wavelengths. In fluorescence spectroscopy the interest lies in measuring the fluorescent light from the sample. This is done by radiating the sample with light at a specific wavelength and then collecting the light emitted from the sample. The collected light will contain both fluorescent wavelengths and incident wavelengths due to light scattering. Longer fluorescent wavelengths are separated from excitation wavelengths and its intensities are measured.

Figure 5, Fluorescence spectrophotometer design

Excitation wavelengths are typically singled out from a broad spectrum source, e.g. xenon arc or mercury vapor lamp, using a filter or a monochromator setup. A common design is to place the detector at a 90º angle to the excitation light to reduce excitation wavelengths and increase the signal-to-noise ratio (see figure)

Fluorescence spectroscopy analysis was performed with an ISA Jobin Yvon-Spex FluoroMax-2 fluorescence spectrometer using photo-excitation at 400 nm with the sample in 1 cm plastic cuvettes and the Hitachi FL F4500 also using photo-excitation at 400 nm.

12

3.6 Fluorescence microscopy

The basic function of a fluorescence microscope is very similar to the fluorescence spectrophotometer. First the specimen is radiated with a specific band of wavelengths and then the much weaker emitted fluorescence is separated from the excitation light. Fluorescent light is transmitted to an ocular or in many cases a digital camera. The light source used in fluorescence microscopy is typically a broad wavelength emitting lamp and excitation filters are used to exclude all but the desired wavelengths. Most fluorescence microscopes are epi-fluorescence microscopes, i.e. excitation and observation of the fluorescence are from above (epi) the specimen.

Figure 6, epi-fluorescence microscope

13

The investigation of surface immobilised polyelectrolytes, biomolecules and the complex between them on patterned surfaces was performed using the Zeiss Axiovert epi-fluorescence microscope A200 Mot equipped with an Hg lamp (HBO 100) as the light source. The microscope has four different filters; 365/12 nm (397 nm long pass emission filter (LP)), 405/30 nm (450 nm LP), 470/40 nm (515 nm LP) and 546/12 nm (590 nm LP), which were used with various exposure times in different experiments. The microscope was equipped with an AxioCam HRc CCD camera. Images of the experiments were analysed using the Zeiss AxioVision PC software v3.1 program. Standard silicon wafers can not be used for fluorescence microscopy as the electrically insulating silicon dioxide layer is not thick enough to prevent luminescence quenching of the fluorescent molecules[18].

3.7 Imaging ellipsometry

Ellipsometry is a very sensitive optical method for determining characteristics of surfaces. By measuring the change in polarization state of a light beam reflected from a surface, information about the surface can be deduced.

The name “ellipsometry” is indeed derived from “ellipse”, in this case referring to the polarization state of the electric field of light. Monochromatic light may at any point in space be resolved into three independent harmonic oscillations along an x,y,z-coordinate system. Let the light wave propagate as a plane wave along the z-axis. The vector of the electric field, E, is then always orthogonal to z and can be described as two harmonic, same frequency oscillations along x and y. These oscillations are generally of different amplitude and phase, thus the E vector moves along an ellipse when the plane wave passes a plane in space. The behavior of an electric field with time at fixed plane in space is called polarization. If the two x and y oscillations of the electric field are of the same phase the vector will travel along a line and is said to be linearly polarized, if the oscillations are exactly 90 degrees out of sync the ellipse will become a circle. Thus linear and circular polarizations are specialized cases of elliptical polarization.

The polarization state of light incident upon a surface is typically decomposed into an s and a p component (the s-component is oscillating normal to and the p-component is oscillating parallel to the plane of incidence). The intensity of the s and p components, after reflection, is denoted by Rs and Rp. Upon reflection on a substrate with a transparent isotropic thin film, the intensity and the phase, δ, of the two components changes. These changes are not the same for both components and it is these differences that are measured and used to quantify the surface. The fundamental equation of ellipsometry is given by[19]

∆−⋅Ψ=⋅=

ℜ

ℜ≡ ii

s

p

s

pee

R

Rsp tan

)( δδρ

A common type of ellipsometer is the polarizer- compensator-sample-analyzer (PCSA) type, the name referring to the order of the different components in the instrument’s light path.

14

1) A monochromatic and collimated light beam is first linearly polarized by the polarizer (P). 2) When passing through the compensator (C) a known phase shift between the x- and y- component of the electric field is induced. 3) Reflection from the sample (S) causes an additional phase shift (∆) and a relative amplitude change (tan Ψ). 4) The light passes another linear polarizer called analyzer (A). Finally a detector measures the intensity transmission through the whole system. The azimuth of the different components can be controlled to a very high precision.

Figure 7, Schematic of PCSA type ellipsometer

As we are looking for complete extinction of the light (nulling) at the detector the strategy is to adjust the different optical components until this condition is met. Extinction can only be accomplished if the light is linear polarized before passing the analyzer and totally blocked out from reaching the detector by the analyzer. As the compensator typically induces elliptical polarization the only way to generate linear polarization is if reflection from the sample reverses the phase shift.

Using this setup it is possible to derive an expression for the intensity variation with the azimuth angles P, C and A.

[ ] [ ])sin(cos)cos(sinsin)sin(sin)cos(coscos CPCiCPCARCPCiCPCARE sp

t

A −−−+−+−=

As we are trying to find the condition when the flux falling on the detector is zero the expression above can be written as

)tan(1

)tan(tantan

CPi

CPiCA

R

R

s

p

−+

−−−=≡ρ

It is customary to fix the compensator azimuth to 45degrees and we have

)45(2tan)45tan(1

)45tan(0tan °+=

°++

°++= PiAe

Pi

PiAρ

Remembering the initial definition we see that

°+=∆

=Ψ

902P

A

15

The complex reflection coefficients are given by

β

β

201

21201

1 i

p

i

pp

per

errR

−

−

+

+=

β

β

201

21201

1 i

s

i

sss

er

errR

−

−

+

+=

β

β

β

β

ρ2

1201

21201

21201

21201 1

1 i

ss

i

ss

i

pp

i

sp

s

p

err

err

err

err

R

R−

−

−

−

+

+⋅

+

+==

Where mnsr and mnpr are the Fresnel coefficients for s- and p-polarized light, respectively,

the subscripts 01 (mn) and 02 denote the ambient/film interface and film/substrate, respectively

nmmn

nmnmmnp

NN

NNr

φφ

φφ

coscos

coscos

+

−=

nnmm

nnmmmns

NN

NNr

φφ

φφ

coscos

coscos

+

−=

The angle of transmission is given by Snell’s law

2211 sinsin φφ NN =

β is called the film phase thickness. Given by

11 cos2

φλ

πβ N

d⋅=

Figure 8, 3-phase ellipsometry

with =1N complex refractive index of the film, =1φ refraction angle in the film,

=λ wavelength of the light and =d thickness of the film. This gives a function

16

),,,,,( 0210 λφρ dNNN dependent on nine real-valued arguments, the real and imaginary

part of the three complex refractive indices, 210 ,, NNN , the film thickness d, the angle of

incidence, 0φ and the wavelength, λ . It is generally not possible to analytically invert

one of the unknown parameters. Many numerical computer inversion algorithms have been developed for this. However, measurement of the complex reflectance ratio at a specific angle of incidence and wavelength provides enough information to determine two real value parameters, assuming that all the remaining parameters are known. If only one parameter is unknown for example the thickness d, the problem is overdetermined and an exact solution can not be found in general.

Imaging ellipsometry is a development of ellipsometry. By adding an objective and a spatially resolving detector e.g. a sensitive CCD camera one can get a detailed image of the sample. As the laser beam is scanned over the sample surface the objective projects the illuminated area onto the camera. Depending on the optical characteristics of the specific area, the camera signal will differ correspondingly. If the reflected light is extinguished, “nulled”, by the particular P, C A-settings the image will appear dark. Regions not fulfilling the null-condition appear brighter, it is now possible to change of P, C, and A it is now possible to find the Null for these regions, which will cause the former dark areas to appear bright. Applying proprietary algorithms allows one to map the Nulls for the entire image, if necessary. This yields a two-dimensional map of the ellipsometric data that can be transformed into a thickness map of the sample.

The thicknesses reported when using an ellipsometer are clearly heavily dependant upon the actual optical model and numerical solution algorithms used. This adds a factor of uncertainty one needs to be aware of. Supporting thickness measurements with other methods is always encouraged.

All ellipsometry measurements in this work were performed with the imaging null ellipsometer EP3 (Nanofilm Surface analysis, Germany), using a laser wavelength of 532.0 nm. The software EP3 View V2.05 was used to calculate the thickness maps. The mediums were all regarded as non-absorbing i.e. the imaginary part of the refractive indices were set to zero. The real parts of the refractive indices in the optical model are listed in table 1. Silicon wafers with a thicker ~1000 Å silicon dioxide layer or transparent glass substrates can not be imagined using these models.

Table 1, refractive indices used in ellipsometry

Medium Real part of refractive index

Air 1.00

Printed biomolecules 1.50

Silicon oxide 1.46

17

3.8 AFM

Atomic force microscopy is a powerful tool for topographical evaluation, presented by Binning et al in 1986, based upon scanning tunneling microscopy. Its main component is a millimeter-long cantilever with a very sharp tip protruding from one end. The cantilever is typically composed of silicon or siliconnitride and the tip size is in the order of nanometers. The tip is brought into close proximity to a sample surface. The van der Waals forces between the tip and the sample will deflect the cantilever according to Hookes law klF ⋅∆−= , where the spring constant, k, of the cantilever is known and the deflection is measurable. Scanning the cantilever, connected to a feedback system, over a surface gives force information if the height over the sample is kept constant or topographic information if the force is kept constant.

In practice the sample is moving under the probe, not vice versa. Generally the sample is mounted on to and controlled in x and y directions, for scanning, and z direction, for maintaining force or height, by a piezoelectric tube.

The primary modes of operation are constant contact mode, non contact mode and dynamic contact mode. In constant contact mode the tip scans very close to the sample surface and the forces acting upon it are repulsive. Very much like a stylus profilometer. The deflection of the cantilever is measured and compared in a DC feedback amplifier to some desired value. A feedback mechanism corrects the deflection by raising or lowering the sample. The main problem with this mode is that the repulsive force can not be kept small enough. As the tip is scanned along the surface this leads to disruptive friction forces.

In the non contact modes the tip is scanned some hundred Ångström above the surface. Van der Waals forces attract the tip and deflect the cantilever. In this mode the tip is ideally never touching the surface and thereby destroying the sample. As these forces are much weaker than the ones involved in contact mode, the sensitivity of measuring changes in a dynamic process is utilized. The cantilever is made to oscillate at or close to its resonance frequency; the attractive forces change the amplitude, frequency and phase of the oscillation. This change can be measured with high precision.

18

Figure 9, Schematic of AFM setup

Dynamic contact mode is a combination of the aforementioned techniques. The cantilever is made to oscillate and the tip is let to come in contact with the surface every cycle, enough force is applied to make the tip detach. The change in oscillation amplitude of the cantilever relative to the free air amplitude can be used to create a topographical map of the sample. This mode is more commonly known as tapping mode. The phase lag of the cantilever oscillation, relative to the driver piezo oscillation is also a useful source of information. A phase map can reveal information about variations in composition, adhesion, friction and viscoelasticity of the sample surface. This makes it possible to use phase information to characterize a surface lacking differences in height of adsorbed species but with different chemical properties. The lateral resolution of the instrument is highly dependant of the tip quality and shape. Optimal vertical resolution can be as high as 0.5 nm.

Atomic force microscopy (AFM) imaging was performed in tapping mode on a SFM-Nanoscope III, Digital Instruments. Cantilevers for tapping mode were obtained from NT-MDT. AFM was used to evaluate printed patterns on silicon substrates. Glass substrates are not suitable for AFM imaging due to the high surface roughness

19

4 Results and discussions

4.1 Fluorescence spectroscopy

4.1.1 Comparison of protein interaction between four different conjugated polyelectrolytes

Spectroscopic investigations of four different conjugated polyelectrolytes (POWT, POMT, PTT, PTAA) interacting with human serum albumine (HSA) and anti-human serum albumine (aHSA) in solution were carried out. The aim of these experiments was to find a suitable candidate for further investigations. A fluorescence spectrum of the polyelectrolyte (0.8 µM) in phosphate buffer (pH 7.4) was recorded. Antibodies were added to the solution (to a final concentration of 0.8 µM CPE, 0.8 µM aHSA) and another spectrum was recorded. Lastly HSA was added to the polyelelectrolyte- HSA solution (0.8 µM CPE, 0.8 µM aHSA, 1.0 µM HSA) and a final spectrum recorded. The wavelength shift of the fluorescent peaks when HSA was added was compared between the four conjugated polyelectrolytes. Gaussian multi-peak fits were made with the software OriginPro 7.5 (OriginLab Corp.,Northampton, MA, USA)

4.1.1.1 Spectrum analysis of POWT

450 500 550 600 650 700

0

100000

200000

300000

400000

500000

600000

700000

800000

POWT-aHSA POWT-aHSA-HSA0,0

0,5

1,0

1,5

Flu

ore

scence (

a.u

)

Wavelength (nm)

POWT

POWT-aHSA

POWT-aHSA-HSA

Gaussian fits

Peak markers

Heig

th r

atio 5

40/5

80

Polyelectrolyte

POWT-aHSA POWT-aHSA

-HSA

Figure 10, fluorescence spectrum (ex. 400 nm) of POWT interacting with HSA and aHSA

20

Applying a 4-peak Gaussian fit to the fluorescent data show two interesting peaks at 540 nm and 580 nm. Upon adding aHSA, fluorescence activity is heightened overall. When HSA is added to the polyelectrolyte-antibody solution the intensity ratio between the two fluorescent peaks is clearly altered but the intensity is not increasing.

4.1.1.2 Spectrum analysis of POMT

450 500 550 600 650 700

0

100000

200000

300000

400000

500000

600000

POMT-aHSAPOMT-aHSA-HSA0

0,1

0,2

0,3

0,4

0,5

0,6

Flu

ore

sce

nce

(a

.u)

Wavelength (nm)

POMT

POMT-aHSA

POMT-aHSA-HSA

Gaussian fits

Peak markers

POMT-aHSA POMT-aHSA

-HSA

He

igth

ra

tio

565

/60

0

Polyelectrolyte

Figure 11, fluorescence spectrum (ex. 400 nm) of POMT interacting with HSA and aHSA

Applying a 4-peak Gaussian fit to the fluorescent data show two interesting peaks at 565 nm and 600 nm. Upon adding aHSA, and subsequently HSA, fluorescence activity is lowered overall. When HSA is added to the polyelectrolyte-antibody solution the intensity ratio between the two fluorescent peaks remain the same.

21

4.1.1.3 Spectrum analysis of PTT

450 500 550 600 650 700

0

100000

200000

300000

PTT-aHSA PTT-aHSA-HSA0,0

0,5

1,0

1,5

Flu

ore

sce

nce

(a

.u)

Wavelength (nm)

PTT

PTT-aHSA

PTT-aHSA-HSA

Gauss fits

Peak markers

Heig

th r

atio 5

60/6

15

Polyelectrolyte

PTT-aHSA PTT-aHSA

-HSA

Figure 12, fluorescence spectrum (ex. 400 nm) of PTT interacting with HSA and sHSA

A 3-peak Gaussian fit of the fluorescent data show two possible interesting peaks at 560 nm and 615 nm. Upon adding aHSA fluorescence activity is lowered overall. When HSA is added to the polyelectrolyte-antibody solution the intensity ratio between the two fluorescent peaks is slightly altered.

22

4.1.1.4 Spectrum analysis of PTAA

425 450450 475 500500 525 550550 575 600600 625 650650 675 700700 725

-100000

0

100000

200000

300000

400000

500000

600000

700000

800000

PTAA-aHSA PTAA-aHSA-HSA0,0

0,5

1,0

1,5

Flu

ore

sce

nce (

a.u

)

Wavelength (nm)

PTAA

PTAA-aHSA-HSA

PTAA-aHSA

Gaussian fits

Peak markers

Heig

th r

atio

535/5

85

Polyelectrolyte

PTAA-aHSA PTAA-aHSA

-HSA

Figure 13, fluorescence spectrum (ex. 400 nm) of PTAA interacting with HSA and aHSA

A 3 -peak Gaussian of to the fluorescent data reveal two peaks at 535 nm and 585 nm. Upon adding aHSA, and subsequently HSA, fluorescence activity is heightened overall. When HSA is added to the polyelectrolyte-antibody solution the intensity ratio between the two fluorescent peaks remain is somewhat changed.

4.1.2 Interaction of POWT and aHSA

The stoichiometric ratio between polyelectrolyte and antibody in solution is of relevance to the fluorescence. At an optimal ratio there would be no unassociated POWT-molecules and concurrently no IgG-molecules without associated POWT-chains. Antibodies without reporter molecules will lower the sensitivity of the sensor systems as well as be economically unwise. If there are many unassociated POWT-chains in the solution there is a tendency for them to aggregate together in larger conjugative systems (inter-chain event). This shifts the fluorescence towards longer wavelengths. By increasing the aHSA-to-POWT ratio and monitoring the intensity ratio associated with inter-chain and intra-chain event it is possible to find a stoichiometric ratio where adding more POWT will not change the intensity ratio towards longer wavelengths. The optimal stoichiometric is then just before the plateau is reached.

A series of fluorescence measurements with increasing aHSA-to-POWT ratio was carried out. The results are presented in Figure 34. The previously identified peaks at 580 nm and 540 nm were analyzed further by plotting the concentration ratio versus fluorescence intensity ratio at the aforementioned wavelengths. See Figure 14

23

-0,2 0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6

0,75

0,80

0,85

0,90

0,95

1,00

Flu

ore

scence r

atio

(5

40nm

/580nm

)

Concentration ratio (POWT/aHSA)

Figure 14, concentration ratio versus fluorescence ratio of POWT-aHSA

4.2 Fluorescence spectroscopy – discussion

Upon adding biomolecules to a polyelectrolyte solution the fluorescence intensity is generally heightened over the whole visible spectrum. This is true both for adding antibody and subsequently corresponding antigen. A possible explanation for this is less aggregation of the polyelectrolyte chains as they associate to the added biomolecules. This have been shown earlier by Nilson et. al. with CPEs associated to DNA. [17] One curious exception to this tendency is solutions of POMT. When biomolecules are added the overall fluorescence is lowered overall. The first attempt to explain this would be to redo the experiment and make sure that no mistake was done when labeling the data. Another possibility is that polyelectrolytes and biomolecules form heavy aggregates that precipitate thus lowering fluorescence intensity. This is in contrast to the results from PTAA spectrometry, fluorescence intensity is increased both when aHSA and HSA is added to the solution.

24

A major shift in peak intensity fluorescence is an advantage when using conjugational polyelectrolytes as part of sensing devices. Of the four CPEs tested POWT gives rise to the largest shift, in fact an increase close to 80% of the original ratio. This is compared to a 10% increase using PTT and PTAA and no increase when using POMT. The intensity increase from PTAA-aHSA complex when adding HSA might be in its favor when used as a reporter molecule. The increase can probably not be used as the sole signal though as chain separation can result from many factors other than complement binding.

Results from concentration titration experiments of POWT versus aHSA are somewhat unsatisfying. Other work has shown a similar titration curve including the redshift “dip” at concentration ratio 0,1 [20] and an planar plateau at high concentrations. The last (highest concentration ratio) titration point is odd. Perhaps increasing concentrations would have revealed it as an outlier. It is difficult to draw unambiguous conclusions from this data but an educated guess is that an possible optimal concentration ratio is somewhere between 0,4 and 0,8 i.e. between 1 and 3 POWT chains per IgG molecule. Considering the relative size of the molecules (3 kD versus 150 kD, POWT chains can be described as “tagging along” the antibody.

4.3 Microcontact printing

4.3.1 Microcontact printing of antibody

Studies were done to evaluate the printing process described in previous chapters. The species printed were in most cases Texas red conjugated aStreptavidin. It is believed that all fluorescent molecules are conjugated to antibody molecules i.e. all fluorescence in the patterns can be associated to antibody molecules. A layer of antibody molecules protruding from the surface are expected to be thicker than a corresponding layer oriented parallel to the surface. See Figure 15

Figure 15, sketch of antibodies adsorbed to a substrate [21]

Different studies aiming to investigate different properties of microcontact printing were performed. The printing protocol was sometimes modified between the studies in order to find certain relationships between the method of printing and the properties of the final print. The printing protocol is outlined below.

Substrate preparation. The substrates are cleaned of organic material.

Inking of stamps. Rubber stamps are exposed to the material to be printed.

Drying of stamps. Fluid are removed from the stamp in order to make clear prints

25

Printing. The stamp is put in contact with the substrate.

Rinsing of prints. Superfluous and non-adsorbed material is removed from the substrate.

Substrate incubation. The prints are exposed to a solution containing substances of interest.

4.3.1.1 Effect of moisture on stamp prior to printing

The effect of introducing a thin layer of condensed water on the stamp surface before printing has been discussed in literature [25]. In order to compare this practice with a printing procedure without this condensation step a series of contact printing experiments were conducted. PDMS stamps were treated in an air-plasma chamber for 12 seconds. The stamps were inked for 10 minutes in a solution of aHSA (25 µg/ml) and POWT (5 µg/ml) in PB (20mM, pH 7.4). After being completely blown-dried with N2 the stamp was cooled in a cabinet freezer (-17°C) for 1 minute and then ambient moisture were let to condensate on the stamps for 10 seconds. A control group of stamps were left in ambient moisture for 2 minutes before printing. Both types of stamps show similar printing results, examined with the fluorescence microscope. Therefore in later experiments, stamps were not cooled before printing.

No studies were performed in order to evaluate the relative bioactivity of the prints produced with the two different methods.

4.3.1.2 Effect of pH and stamp surface energy on printed antibody

patterns

In order to test the relationship between inking solution pH and quality of printed patterns four samples were examined using imaging ellipsometry. Oxygen plasma treated PDMS stamps were incubated with protein solution at pH 5.5, pH6.8, pH 7.4 and pH 9.0 respectively. Stamps were then rinsed with PB at respective pH and then put in contact with TL-1 washed silicone oxide substrates and again rinsed multiple times with buffer adjusted to the four different pH levels.

Table 2, Microcontact printed Streptavidin-TxR printed on silicon using hydrophilic stamps.

Estimations of pattern thickness made from ellipsometric surface maps.

Sample Median height

Hydrophilic stamp, pH 5.5 5 nm

Hydrophilic stamp, pH 6.8 4 nm

Hydrophilic stamp, pH 7.4 1 nm

Hydrophilic stamp, pH 9.0 1 nm

It is believed that IgG molecules are adsorbed to the stamp and transferred to the substrate at all the four pH-levels tested.

26

Prints on glass substrates examined in fluorescence microscopy show spots of high local fluorescence intensity, referred to as aggregates. This suggests high local concentration of antibody. Aggregation spots seem to be of higher abundance at lower pH levels (see Figure 17), while printing results from stamps inked in pH 7.5 and pH 9.0 antibody solutions are more homogeneous. This could be accounted to permanent charge acquisition or denaturation of the IgG molecules due to the low pH. IgG molecules in general have an isoelectric point pH 6.1-8.5 [23]

Another set of Streptavidin-TxR prints on silicon oxide using native hydrophobic rubber stamps and four different pH levels (pH 5.5, pH 6.8, pH 7.4 and pH 9.0) were evaluated with imaging ellipsometry and AFM.

Figure 16, aStreptavidin-TxR on silicon (pH 7.4)

Ellipsometry results from these experiments are not easily interpreted (all results not shown). Thicknesses of printed antibody layers are below 2 nm indicating low surface coverage. The patterns were in general of poor quality. However, some measurements show thicker antibody layers (3-5 nm) in well defined patterns. (See Table 3 and Figure 16)

Table 3, aStreptavidin-TxR on printed on silicon using hydrophobic stamps. Measurements

of pattern thickness were made with individual one zone measurements.

Imaging ellipsometry measurements were not enough to fully evaluate the impact of pH and stamp surface energy on printing quality. No clear decision regarding preferable pH could be made based on these results; however the most satisfactory results were achieved at a close to pH-neutral buffer. Thus PBS (20 mM) pH 7.4 buffer solution was used in later experiments. The slightly higher thickness of the samples printed with a hydrophobic stamp suggest that the printed species are more densely adsorbed on the surface.

Sample mean

thickness (nm)

standard deviation

(nm)

number of

measurements

Hydrophobic stamp 1, pH 7.4

3.6 1.4 6

Hydrophobic stamp 2, pH 7.4

5.8 2.1 11

27

Figure 17, aStreptavidin-TxR on glass, pH 5.5,

excited at 546 nm, 1800 ms exposure time

Figure 18, aGABA-POWT printed on glass,

excited at 470 nm, 10s exposure time.

Figure 19, aStreptavidin-TxR on glass incubated

with Streptavidin-FITC, excited at 546 nm, 4300

ms exposure time

Figure 20, aStreptavidin-TxR on glass incubated

with Streptavidin-FITC, excited at 470 nm, 6800

ms exposure time

Figure 21, Streptavidin-FITC printed on glass,

incubated with aStreptavidin-TxR, excited at 470

nm

Figure 22, Streptavidin-FITC printed on glass,

incubated with aStreptavidin-TxR, excited at 546

nm

28

4.3.1.3 AFM evaluation of microcontact prints

AFM measurements of the same samples were made to compare the results from imaging ellipsometry (see Figure 16) and to obtain more detailed information of the printed antibody layer.

Figure 23, AFM scan of aStreptavidin-TxR prints on silicon,

pH 7.4, shown is the circumference of a printed disc

50 µm diameter, 6 µm scan size

Figure 24, AFM scan of aStreptavidin-TxR prints on silicon,

pH 7.4, shown is the circumference of a printed disc

50 µm diameter, 50 µm scan size

29

This resolution reveals that the molecular species printed is arranged in individual clusters, hundreds of nanometer in size and clearly spaced. Considering the network pattern revealed by AFM it is likely that the transferred material is indeed IgG-molecules [24]. AFM measurements show the height of the printed molecules to be 10-15 nm. This is significantly higher than previously measured by ellipsometry. A possible explanation to this is that the laser beam used in imaging ellipsometry is much wider than the AFM tip. The results reported by the ellipsometer are thus an average thickness while the very thin AFM tip can make a distinction between high and low points on the sample surface. Worth noting is also the sharp contrast between printed and non printed areas. A cluster of high structures can be seen in the top left corner of Figure 24 and also in the center of the “dots” in Figure 16. This can be a sign of aggregates of antibodies as seen in Figure 17.

4.3.2 Microcontact printed antibody in complex with CPE

As POWT was shown to have the largest fluorescence shift when adding antigen to POWT/antibody solution it was chosen for microcontact printing experiments. GABA is one of the antigens that might be interesting to detect in a later application of this technique. aGABA was mixed with POWT in PBS (pH 7.4) in a 0.8 ratio (0.8 µm aGABA, 1.0 µm POWT) in order to create an antibody-CPE complex. The solution was inked on a hydrophobic PDMS stamp for 60 minutes and let in contact with a TL1-washed glass substrate for 30 minutes

Figure 18 show fluorescing patterns of microcontact printed POWT. Because of the interaction between POWT and IgG molecules showed earlier it is assumed that antibodies are present in the printed pattern.

Subsequent incubation of antigen (GABA) did not induce a change in color detectable by visual inspection, however.

4.3.3 Microcontact printed antibody interacting with antigen

PDMS stamps were treated in air plasma to produce a hydrophilic surface. They were then inked with aStreptavidin-TxR solution (50 µg/ml) for 60 minutes. The stamps were then put in contact with glass substrates for 14 hours. The patterns were incubated in HSA solution (1%) for 45 minutes before incubated with Streptavidin-FITC for 2 hours.

Red fluorescence between areas of contact (100 µm wide bands are non-contact areas, see Figure 19 and Figure 20) indicates that aStreptavidin-TxR is not confined to areas of contact. Two possible explanations are proposed.

1. Incompletely cured siloxanes are known to diffuse from PDMS/substrate interfaces. This phenomenon might spread fluorescent material from the stamp over the surface.

2. Fluorescent material migration over the surface is facilitated by solvent in later incubation or washing steps.

An explanation to the very irregular antibody patterns produced in this experiment and especially the lack of fluorescence in expected contact areas might be related to the long contact time. PDMS put in contact with glass will adhere to the surface over time due to the migration of siloxane monomers. This process produces local encapsulation of fluorescent material which then stays on the stamp when it is removed.

30

Since aSteptavidin-TxR has been shown to fluoresce in the green wavelength region it is not possible to differentiate this contribution to the green light intensity from the potential fluorescence from the antigen conjugated FITC. Even though the substrates were incubated with HSA-solution substantial amounts of Streptavidin-FITC is adsorbed to the surface during the second incubation.

4.3.4 Microcontact printed antigen interacting with antibody

Hydrophobic PDMS-stamps were incubated with Streptavidin-FITC for 60 minutes, rinsed with PB (pH 7.4) and printed on TL-1 washed glass substrates. After the print was incubated with aStreptavidin-TxR solution for 90 minutes a pattern was seen in the fluorescence microscope. (See Figure 21 and Figure 22)

Results presented can be interpreted as there is an interaction between printed antibody and subsequently incubated antigen. Printing of antigen and subsequent incubation with fluorescent labeled antibody clearly indicates the specificity of the interaction. However, without control prints using bare stamps it’s not possible to rule out that residues from the stamp have an affect on antibody adhesion.

4.3.5 Printed antibody interacting with antibodies in solution

4.3.5.1 Printing protocol

Glass object slides were cleaned in NH3-acid (65%) for 24h and then rinsed in flowing deionized water for 20 minutes. They were kept submersed in water until dried with N2-gas and air plasma cleaned prior to printing. PDMS-stamps consisting of lines 100µm wide with increasing spacing were molded using 10:1 base-curing mix, degassed, cured for 1h at 120 degrees Celsius. The stamps were then ultrasonically cleaned in 50% EtOH for 10 minutes and then rinsed in multiple baths of deionized water. Stamps were dried with flowing N2-gas before inking. Solutions of Rabbit aStreptavidin-TxR (50µg/ml in PB, pH 7.5), aRabbit-Alexa (50µg/ml in PB, pH 7.5) and Streptavidin-FITC (50 and 25 µg/ml in PB, pH 7.5) were prepared. 50 µl of primary inking solution was applied to the stamp surface and kept dark for 60 minutes. Two samples were bath incubated in primary solution for reference. The protein inked stamps were then rinsed in 3 sequential PBS-baths (pH 7.4) and dried with flowing N2-gas. Stamps were put in contact with the glass cover slips for 60 minutes. Substrates were incubated with secondary protein solution for 60 minutes. All substrates were then washed with PBS to remove weakly bound biomolecules. A final wash with deionized water was done to remove salt residues prior to drying with N2 flow and microscopy.

31

Table 4, experiment outline

Sample Primary solution Secondary solution

1 aRabbit-Alexa Rabbit aStreptavidin-TxR

2 aRabbit-Alexa Rabbit aStreptavidin-TxR

3 aRabbit-Alexa Streptavidin-FITC

4 aRabbit-Alexa Streptavidin-FITC

5 Rabbit aStreptavidin-TxR aRabbit-Alexa

6 Rabbit aStreptavidin-TxR aRabbit-Alexa

7 Rabbit aStreptavidin-TxR (bath incubated)

Streptavidin-FITC

8 aRabbit-Alexa (bath incubated)

Rabbit aStreptavidin-TxR

4.3.5.2 Printed antibody interacting with antibodies in solution – results

Incubated substrates were investigated in the Zeiss Axiovert A200 fluorescence microscope.

32

Figure 27, Sample 3 excited at 365 nm, 1000 ms

exposure time

Figure 28, Sample 3 excited at 470 nm, 1000 ms

exposure time

Figure 29, Sample 5 excited at 546 nm, 1000 ms

exposure time

Figure 30, Sample 5 excited at 365 nm, 1000 ms

exposure time

Figure 25, Sample 2 excited at 365 nm, 1000 ms

exposure time Figure 26, Sample 2 excited at 546 nm, 1000 ms

exposure time

33

Figure 31, Sample 7 excited at 365 nm, 1000 ms

exposure time

Figure 32, Sample 7 excited at 365 nm, 1000 ms

exposure time

Figure 33, Sample 8 excited at 546 nm, 1000 ms

exposure time

500 550 600 650 700

0

100

200

300

400

500

Flu

ore

sce

nce

(a

.u.)

Wavelength (nm)

no aHSA

48

24

18

12

6

3

2

1.5

1

0.6

0.4

stoichiometric ratio

POWT/aHSA

Figure 34, fluorescence spectrum of POWT-aHSA complex at increasing ratios. Curves

shifted 100 a.u on y-axis for clarity

34

4.3.5.2.1 Printed aRabbit-Alexa incubated with Rabbit aStreptavidin-TxR

Micrographs above (see Figure 25 and Figure 26) show examples of pattern details where there are no visual signs of printed primary antibody and still high presence of secondary antibody. In areas where the stamp, due to pattern design or stamp defects, has not been in contact with the substrate neither the primary nor the secondary antibody was detected. This indicates that changes of the substrate surface due to intrinsic properties of the stamp are more important for secondary antibody adsorption than contact printing of primary antibody. Previous work has shown that PDMS-stamps can leave residues on glass substrates and thus increase the surface energy of the substrate leading to increased IgG-molecule adsorption [22]

4.3.5.2.2 Printed aRabbit-Alexa incubated with Streptavidin-FITC

Figure 27 reveal a printed pattern of Alexa marked IgG but this pattern can be seen in Figure 28 as well indicating that Streptavidin-FITC has adsorbed to the printed surfaces. As there is no biological specificity between Streptavidin-FITC and IgG-Alexa) the driving force for Streptavidin-FITC adsorption must be something else.

4.3.5.2.3 Printed Rabbit aStreptavidin-TxR incubated with aRabbit-Alexa

The prints of Texas red conjugated antibodies are very clear and it is easy to verify the existence of printed antibodies. The micrograph (see Figure 29) reveals some defects in the printed pattern. There are no clear signs of specific binding of aRabbit-Alexa to the printed antibodies (see Figure 30). However, the general low fluorescent intensities from Alexa make evaluation difficult.

4.3.5.2.4 Bath incubated aRabbit-Alexa incubated with Streptavidin-FITC

Micrographs of sample 7 show very low fluorescent intensities and naturally no patterns. It is interesting to compare Figure 32 with Figure 28. The fluorescence seen in Figure 28 is nowhere to be seen in Figure 32, indicating that the printing process itself is enough to create a discriminating pattern for adsorbtion. The mechanism for this is most probably surface energy modification.

4.3.5.2.5 Bath incubated aRabbit-Alexa incubated with Rabbit aStreptavidin-

TxR

When comparing Figure 33 with Figure 25 it’s easy to see that the printing process clearly affects the adsorption patterns. More aStreptavidin-TxR is adsorbed to the printed substrate (Figure 25) than if bath incubated (Figure 33), regardless of printed species.

35

4.4 Microcontact printing – discussion

The printing process of fluorescent material was tested and the production protocol is robust. With quality rubberstamps high resolution patterns can be produced with relative ease. Fluorescence microscopy, force microscopy and ellipsometry give coinciding results regarding the size and high contrast of the printed patterns.

The major driving force behind secondary antibodies adhering to substrates in patterns shown is probably the change in surface characteristics induced by residues from the PDMS stamp.

Since the substrates used in imaging ellipsometry can not be used in fluorescence microscopy (and vice versa) it has not been confirmed by fluorescence microscopy that antibodies conjugated with Texas Red have been printed to silicon substrates. It’s most likely however that the pattern seen in ellipsometry and AFM is printed antibodies.

Results from fluorescence measurements show that CPE interacts with antibodies in solution. Using mixtures of antibodies and CPE in inking solutions makes it plausible that antibody molecules are localized to fluorescent printed areas. It could not be shown however that antigen in solution interacted with immobilized antibody-CPE-complex. In hindsight it’s not certain that the shift seen in spectroscopy is detectable for the human eye. The spectrum from the patterns needs to be analyzed in order to detect the shift from antibody-antigen binding.

A great advantage with using bare eye fluorescence microscopy as a tool for evaluation is the appealing pedagogic notion that all one has to look for is different colors. A shift in color represents presence of target molecule. This simple sensor response is ideal for an end user, but in the research processes a quantifiable signal would be appreciated more.

A sandwich assay could easily be used to determine the possible existence of bound antigen species. This is however by no means a novel technique. It has never been the focus of this work to show that molecular species are transferred during microcontact printing, but to investigate whether microcontact printing of antibodies is a promising method for producing sensor arrays. Secondary antibodies could still be used to determine whether antigen is bound to the immobilized antibodies. When established that the printed patterns are able to discriminate antigen in solution, a next step would be to detect the binding event using conjugated polyelectrolytes.

A great source of frustration during this work was the lack of quantitative measurements. It was not enough to use fluorescence microscopy and fluorescent antibodies and antigen to determine the specificity of printed antibody patterns. We need to measure printed antibody affinity to antigen. With a quantitative method we could much more easily determine the effect of different printing parameters.

36

37

5 Conclusions

In a fluorescence spectroscopy experiment it is shown that the spectroscopic properties of the POWT-aHSA molecule is clearly altered when HSA is introduced to the solution.

A titration series of aHSA vs. higher concentrations of POWT was made to calculate an optimal stoichiometric ratio, i.e. the concentration limit where adding more conjugated polyelectrolyte to the antibody will not increase the fluorescence shift. The optimal ratio was shown to be between 0,4 and 0,8.

It has been shown that microcontact prints of POWT and aGABA mixtures fluoresce.

Fluorescence and imaging ellipsometry prove that antibodies are transferred during printing.

Examples of antigen in solution binding to immobilized antibodies have been shown. It’s not conclusive however if this is a sole effect of biological specificity.

38

39

6 Recommendations

Using printed IgG patterns as part of a biological sensor has shown to be very difficult. One part of the problem is that IgG-molecules are very sensitive and can not be expected to always maintain biological functions when exposed to such a rough treatment as microcontact printing. A possible way to avoid this difficulty is to use not entire IgG molecules but only the antigen recognizing part of the molecules called Fab-fragments. It is believed that Fab-fragments are more robust and could maintain biological specificity when microcontact printed. Another alternative could be to use recognition molecules based on recombination of the binding region of Staphylococcal Protein A (SpA), known as Affibodies. These molecules are very robust as the nonbinding fragment is a relative simple three-helix bundle structure compare to the beta-sheet structure of antibodies[26].

One could also try to create a pattern of recognition molecules without microcontact printing of antibodies. Using the high affinity of SpA to the Fc part of an IgG antibody one might create a pattern of antibodies by first immobilizing SpA to a surface with microcontact printing and then incubating the pattern with IgG. Less randomly orientated IgG molecules and less contact print induced denaturation may result in more biological active antibody layers.

It would be very useful if one could use a quantitative method to determine the affinity to antigen of the printed antibodies. This could be done by determining the amount of bound antigen using a secondary fluorescing antibody, much like an ELISA-assay.

More information about immobilized antibodies and CPE complex can be collected by using a wavelength analyzer together with the fluorescence microscope.

40

41

7 Acknowledgements

I foremost want to thank my examiner Olle Inganäs for giving me the opportunity to write this thesis in the group of Biomolecular and organic electronics.

I’d like to thank all the members of this group for their support and suggestions. Of course nothing would have happened without my wonderful supervisors Anna Herland and Jens Wigenius.

I owe my friends Towe and Gunnar Bergström much gratitude for encouraging me to bring this work to conclusion and for always welcoming me in their home.

I also would like to thank my friends Zhang Feng Ling and Du Chun Xia for inspiring me to continue my endeavors into the never ending enterprise of mastering the Chinese language.

42

43

8 References

1. Xia, Y. and G.M. Whitesides, SOFT LITHOGRAPHY. Annual Review of Materials Science, 1998. 28(1): p. 153-184.

2. Bernard, A., et al., Printing Patterns of Proteins. Langmuir, 1998. 14(9): p. 2225-2229.

3. Kumar, A. and G.M. Whitesides, Features of gold having micrometer to centimeter dimensions can be formed through a combination of stamping with an

elastomeric stamp and an alkanethiol ``ink'' followed by chemical etching. Applied Physics Letters, 1993. 63(14): p. 2002-2004.

4. Chee, M., et al., Accessing Genetic Information with High-Density DNA

Arrays10.1126/science.274.5287.610. Science, 1996. 274(5287): p. 610-614. 5. Graber, D.J., et al., Antigen binding specificity of antibodies patterned by

microcontact printing. Langmuir, 2003. 19(13): p. 5431-5434. 6. Howell, S.W., et al., Patterned protein microarrays for bacterial detection.

Langmuir, 2003. 19(2): p. 436-439. 7. Goh, J.B., et al., Diffraction-based assay for detecting multiple analytes.

Analytical and Bioanalytical Chemistry, 2002. 374(1): p. 54-56. 8. Duffy, D.C., et al., Rapid Prototyping of Microfluidic Systems in

Poly(dimethylsiloxane). Anal. Chem., 1998. 70(23): p. 4974-4984. 9. Schmid, H. and B. Michel, Siloxane Polymers for High-Resolution, High-

Accuracy Soft Lithography. Macromolecules, 2000. 33(8): p. 3042-3049. 10. Atkins P, d.P.J., Atkins' physical chemistry. 7 ed. 2002, Oxford. 11. Nilsson, K.P.R., et al., Conjugated polyelectrolytes: Conformation-sensitive

optical probes for detection of arnyloid fibril forrnation. Biochemistry, 2005. 44(10): p. 3718-3724.

12. Wallace, G.G., et al., Conjugated polymers: New materials for photovoltaics. Chemical Invention, 2000. 30(4): p. 14-22.

13. MacDiarmid, A.G., Nobel Lecture: "Synthetic metals": A novel role for organic polymers. Reviews of Modern Physics, 2001. 73(3): p. 701 LP - 712.

14. Nilsson, K.P.R., M.R. Andersson, and O. Inganas, Conformational transitions of a free amino-acid-functionalized polythiophene induced by different buffer

systems. Journal of Physics-Condensed Matter, 2002. 14(42): p. 10011-10020. 15. Improved photoluminescence efficiency of films from conjugated polymers. 85(1-

3): p. 1383-1384. 16. Nilsson, K.P.R. and O. Inganas, Optical emission of a conjugated polyelectrolyte:

Calcium-induced conformational changes in calmodulin and calmodulin-

calcineurin interactions. Macromolecules, 2004. 37(24): p. 9109-9113. 17. Nilsson, K.P.R. and O. Inganas, Chip and solution detection of DNA hybridization

using a luminescent zwitterionic polythiophene derivative. Nature Materials, 2003. 2(6): p. 419-U10.

18. Asberg, P., P. Nilsson, and O. Inganas, Fluorescence quenching and excitation transfer between semiconducting and metallic organic layers. Journal of Applied Physics, 2004. 96(6): p. 3140-3147.

19. Arwin, H., Thin Film Optics - optical response, heterogeneous media,

polarization, reflection, ellipsometry. 5 ed. 2000, Linköping.

44

20. Fransson, S., Surface energy modification for biochip application, in Department of Physics, Chemistry and Biology. 2006, Linköping University: Linköping.

21. LaGraff, J.R. and Q. Chu-LaGraff, Scanning Force Microscopy and Fluorescence

Microscopy of Microcontact Printed Antibodies and Antibody Fragments. Langmuir, 2006. 22(10): p. 4685-4693.

22. Asberg, P., K.P.R. Nilsson, and O. Inganas, Surface Energy Modified Chips for

Detection of Conformational States and Enzymatic Activity in Biomolecules. Langmuir, 2006. 22(5): p. 2205-2211.

23. R.Thorpe, A.J.a. Immunochemistry in Practice. 1988: Blackwell Scientific Publications.

24. Ortega-Vinuesa, J.L., P. Tengvall, and I. Lundstrom, Molecular packing of HSA,

IgG, and fibrinogen adsorbed on silicon by AFM imaging. Thin Solid Films, 1998. 324(1-2): p. 257-273.

25. Feng, J., et al., A novel process for inking the stamp with biomacromolecule solution used in reactive microcontact printing. Colloids and Surfaces B: Biointerfaces, 2004. 36(3-4): p. 177-180.

26. Nord, K., et al., Binding proteins selected from combinatorial libraries of an [alpha]-helical bacterial receptor domain. Nat Biotech, 1996. 15(8): p. 772-777.